This thesis investigates the following objectives: first, it analyses trends in share option compensation for NEDs during the pre–King III period (before they were stopped). The idea is to determine whether the decision to stop them was triggered by a significant increase in their use. The trend analysis is extended to observe changes in the use of share options for NEDs over the full sample period. The intention of this sub–objective is to measure the extent of compliance to King III’s requirement to stop the use of share option compensation for NEDs. Second, the study exploits the natural experiment, presented by King III’s requirement to stop the use of share option compensation for NEDs, to investigate the impact of share option compensation for NEDs on monitoring executives. In addition, the study investigates how institutional and blockholder ownership affect the relationship between share option compensation for NEDs and monitoring (to see whether they are substitutes). Both institutional and blockholder owners consist of heterogeneous categories with different monitoring incentives; hence, a further analysis examines the moderating impact of these different categories of stakeholders. To measure the level of monitoring, the study focuses on two of the biggest agency problems in South Africa: design of CEO compensation and levels of earnings management. The study is based on a sample of 110 non–financial companies (55 in the treatment group and 55 in the control group) listed on the Johannesburg Stock Exchange (JSE), South Africa, over the period 2002–2016. The bulk of the data used was hand–collected from annual reports, the rest was sourced from financial databases such as Bloomberg, Iress and DataStream. The difference– in–difference regression analysis is the main methodology used but for comparison purposes, the study also applies the normal Ordinary Least Squares (OLS) regression and fixed effects model. To control for the endogeneity problem, the study is based on a natural experiment, which is dubbed the ‘gold standard’ for addressing endogeneity problems. Addressing the endogeneity problem is key to satisfactorily settling the debate on the effectiveness of equity–based compensation in mitigating secondary agency problems. The results of the trend analysis show that the growth in share option compensation for NEDs was not statistically significant during the pre–King III period. These results rule out the possibility that King III’s recommendation to stop the use of share option compensation for NEDs was driven by an explosion in their use. As expected, after the introduction of King III, the use of share options declined significantly – an indication that companies largely complied with the requirement to stop the use of share options as compensation for NEDs. However, not all companies are compliant; this is not surprising, as King III was based on the ‘apply or explain’ approach. Regarding the impact of share option compensation on monitoring, the results consistently show that removing share option compensation for NEDs does not weaken monitoring; it either improves monitoring, or it has no effect. Based on these findings, it is not worthwhile, for shareholders, to use share option compensation for NEDs. They come at a cost, they dilute the shareholding structure yet removing them does not weaken monitoring. Overall, the results support King III’s recommendation to stop the use of share option compensation for NEDs. The results also show that the presence of institutional and blockholder ownership does not improve monitoring after the removal of share option compensation. Hence, neither of these two stakeholders are a substitute monitoring mechanism for share option compensation for NEDs. This is inconsistent with the substitution-monitoring hypothesis. These findings persist, even after a sub–sample analysis of the two categories of institutional ownership (monitoring and non– monitoring institutional owners). A further analysis of different categories of blockholder ownership shows that family, pension and foreign blockholder are not a substitute monitoring mechanism for NEDs share option compensation. But the results for government blockholders contradict this; they are a substitute for share option compensation when analysing real–activities manipulation. However, for the rest of the settings they are not a substitute monitoring mechanism. This confirms the view that different blockholders have different incentives to monitor management, which affects organisational outcomes. The study makes the following contributions: (i) It contributes to the literature by addressing the endogeneity problem using a natural experiment. (ii) The study focuses on a unique institutional context, largely ignored by prior studies on this subject. (iii) The study contributes to the crafting of future corporate governance principles in South Africa and the rest of world, specifically on the design of incentive compensation for NEDs. (iv) By investigating the interaction effects of institutional/blockholder ownership and their different categories, the study provides evidence for the substitution-monitoring hypothesis in South Africa. (v) On the use of share option compensation for NEDs, this study contributes to the literature by showing its impact on mitigating agency problems specifically related to the design of CEO pay incentives.

This dissertation investigates the factors which influence unlisted companies’ compliance with International Financial Reporting Standards (IFRS) in South Africa at three levels: the global level, the country level, and the company level. This dissertation also considers whether taking such factors into account in the standard-setting process would lead to improved IFRS compliance. This dissertation applies a multiple case study method followed by a national wide survey. Thus, the data were collected by reading the IFRS, and through questionnaires and interviews. A total of 41 companies responded to the survey while five (5) companies participated in the case study. This dissertation’s main findings show that at company level, governance and financial people working for these companies are significant factors which influence their IFRS compliance. Further, non-auditing and/or non-accounting companies are highly reliant on their external auditors in order to comply with the IFRS. The findings also show that at country level, economic, legal, cultural and professional factors influence these companies’ IFRS compliance. At global level, the findings show that the international recognition of IFRS, transparency, comparability, understandability, foreign operations, and importing and exporting are the factors which influence these companies’ IFRS compliance. There are several contributions that can be attributed to this dissertation. First, this dissertation contributes to the literature by extending the research regarding factors (or some indicators for factors) which influence unlisted companies’ IFRS compliance and these companies’ experience of using IFRS in South Africa. Further, this dissertation suggests a model which explains those indicators which appear to influence unlisted companies’ IFRS use at company level. Second, this dissertation contributes to the literature by testing five theories (these are, decision usefulness theory, agency theory, stewardship theory, stakeholder theory and institutional theory) on unlisted companies. Third, this dissertation’s original contribution is to use a self-prophecy effect to gain a better understanding of unlisted companies’ predicted continuing use of IFRS.

Purpose: This study examines 26 large and established South African multi-asset unit trusts in order to determine their style and asset class exposure over time. The objective is to ascertain whether South African multi-asset fund managers can realise outperformance, that exceeds what can be realised through exposure to representative, investable, style and asset class indices. Such an analysis assists in identifying unit trust manager skill, but a further consideration is how to combine unit trusts in a suitable manner, to this end portfolio construction tools are utilised to meet illustrative client objectives in a multi-asset context. Methodology: This study uses monthly total return time series for several investable style and asset class indices as well as South African multi-asset unit trust monthly total return time series. Where historical data permits, the period under investigation is from 1 January 2003 to 30 June 2018. Style and asset class exposure is determined using the Returns Based Style Analysis (RBSA) of Sharpe (1992) applying a 24-month rolling window approach. Findings: The equity style exposures estimated using RBSA provide evidence that on average the value style was dominant across the multi-asset high equity unit trusts examined. For the multi-asset low equity unit trusts examined the low volatility style was dominant. Moreover, a large proportion of the variability in returns of many multi asset unit trusts, can be explained by exposure to style and asset class indices. Consequently only 3 out of the 15 multi-asset high equity unit trusts analysed, could realise performance in excess of their custom style and asset class benchmark. As only a limited number of these unit trusts could demonstrate superior security selection ability the implication is that many asset managers stand to be disrupted by lower cost products that provide similar style and asset class index exposure. Originality/Value: Much research has been conducted into the style exposures of SA general equity funds. However, to the author’s own knowledge this is the first study to apply RBSA in a performance context to multi-asset unit trusts, under the new ASISA classification standards. The benefits of portfolio construction tools such as portfolio simulation and the ‘Risk Budgeting’ approach are also discussed and applied in a multi-asset context.

This article-based thesis consists of three main papers that examine the use of fair value accounting in banks and how it can influence behaviour with systemic effects; this helps in understanding the role of fair value accounting in the global financial crisis. The examination consisted of two parts. The first part was the investigation of how fair value accounting was actually used by South African banks. The second part was the development of an analytical model that links together fair value accounting, bank capital regulation and economic outcomes. The South African case study was further divided into two parts. In the first part, a comparative design was used to investigate in detail how fair value accounting was implemented by two South African banks and what their motivations were. The second part sought to answer the question: did South African banks pay out higher dividends based on risky fair value accounting gains? The South African evidence indicates that fair value accounting materially impacts the profit and loss and the regulatory capital of banks. This component of regulatory capital proved to be risky. Dangerous pay-outs resulted from the increase in profits and bank assets grew the most during the period of risky capital formation. It was found that the use of a stock-flow consistent model of the economy was a commonality amongst those that predicted the global financial crisis. A stock-flow consistent model was shown to be descriptive of the South African evidence. The model showed fair value accounting to be at the centre of feedback processes that can weaken the banking system during the economic upswing. The study concludes that fair value accounting is central in processes that weaken the banking system during an economic upswing and thus demonstrates why the current call for prudent accounting in banks is justified. The study expands on current literature in a number of ways. It adds to the literature that fair value accounting is procyclical by demonstrating that this effect is not constant throughout the cycle and is more problematic during the upswing; this differs from the usual argument that fair value accounting accelerates the downturn. The South African empirical evidence showed that fair value accounting for available-for-sale assets is not the only avenue for fair value accounting to be dangerous; fair value accounting adjustments through profit and loss should also be monitored. The analytical model as well as the South African empirical evidence contradicts the common argument that the fair value measurement of financial instruments must be pervasive in a bank and banking system to be dangerous. The South African empirical evidence shows that fair value accounting must be considered a possible avenue of earnings or capital management in banks.

Gram-Charlier expansions provide a tractable way of fitting risk-neutral distributions to asset prices. This allows the model to capture skewness, excess kurtosis and higher moments in observed asset returns. Schlogl (2013) proposes a calibration method to ensure the fitted densities are valid and arbitrage free. This method is implemented with standard foreign exchange options and gives an exact fit when enough moments are included in the calibration process. GramCharlier expansions also result in analytic solutions for many exotic option prices through an extremely general framework. This relies on representing an option as a portfolio of the M-binaries defined by Skipper and Buchen (2003). Geometric Asian options are priced using this approach and compared to the corresponding Black-Scholes prices. Numerical examples highlight the effect skewness and excess kurtosis can have on these option prices, particularly for options that are out-the-money. Gram-Charlier distributions are also combined with Monte Carlo simulations to estimate option prices for calls and geometric Asian options. The results show convergence to the analytical solutions for all cases. Additionally, Gram-Charlier estimates for arithmetic Asian options are calculated and compared to Black-Scholes estimates.

There has been a notable decline in the tax morality of South African taxpayers over the years. Tax morality is a term defined as the willingness of individuals to pay tax and comply with tax laws1 . The concerns over the decline in tax morality were raised by the National Treasury when discussing the widening tax gap between the budgeted tax revenue collections and the actual revenue collected (in the 2017 Budget Speech)2 . The importance of tax morality in South Africa cannot be overstated, the very success of South Africa's democracy is dependent on the very functioning of the fiscal citizenship principle, that means the state and the taxpayers must be committed to the operation of the democracy through the social contract, ie the taxpayers must be willing to pay their taxes and the government must provide the services due to the taxpayer, such as healthcare, education services and efficient infrastructure. In the evaluation of South Africa's tax morality, the tax revenue collection process and methodology must be reviewed. For personal income tax, with the exception of PAYE, taxpayers must declare their income earned in order for the income to be assessed. It is for this reason that tax morality should be regularly reviewed, to ensure that taxpayers are motivated to pay the right amount of tax and are compliant with tax laws. Where this is not the case, symptoms of broken fiscal citizenship can include aggressive tax avoidance, base erosion and profit shifting and tax evasion, to an extent that legislation needs to be constantly reviewed in order to protect the South African tax base. To review the tax morality of the Republic, revenue collection statistics such as tax buoyancy are reviewed, revealing that after the 2008/9 global financial crisis, the revenue collection statistics remained buoyant until the 2017/18 period during which time it decreased to 0.91 indicating a threat to the long-term sustainability of the fiscal policy. The tax revenue collections when compared to GDP statistics also signal concern, and when taking into account the rate at which high networth individuals are leaving the country and their reasons, it is clear that the South African government should be applying more focus and resources in improving tax morality. In order to recommend focus areas to improve tax morality, a review of research performed by the OECD globally into the tax morale of individuals and businesses around the world is used. The research focuses on factors influencing tax morale, which is a good indicator of what motivates taxpayers to participate in, and comply with a tax system. An overarching theme for individuals from the identified factors is the perspective of the government taxpayers have. It is this perspective that has been found in the study to influence whether individual taxpayers around the world comply with and participate in the tax laws of the country. For businesses, the most dominant factor appears to be tax certainty and while this is information derived from a general survey and on a limited population, it provides a general overview or a starting point in the improvement of tax morality. With the factors identified, an evaluation of how they impact the South African population and to what extent the findings may be true for the Republic are investigated. The socioeconomic and institutional factors identified in the global study are relatable in the South African context and the results from the survey show parallels between the global population and the South

Spreads between swap legs referencing floating cashflows of different tenors have widened significantly since the global financial crisis of 2008. This frequency basis can be explained by the presence of “roll-over risk”. Defining the roll-over risk state variables in an affine form, this dissertation prices a credit default swap using an “affine transform” methodology. This price is then compared to that obtained from a traditional Monte Carlo simulation approach. The former is shown to produce accurate results with greater computational efficiency, providing a useful way to price complex financial instruments when the state variables are defined in an appropriate form.

The volatility of capital flows and their adverse impact on macroeconomic and financial variables is a major concern to policy makers, resulting in a debate on whether capital controls or financial (capital account) liberalisation is best suited to managing them. This study argues that a better understanding of the underlying process of the foreign capital flows, that is, whether they are a random walk, a persistent, or an anti-persistent series, is a critical but currently lacking element in informing this debate. Specifically for foreign portfolio equity flows, there may also be need to understand their dynamic impact on stock markets. The purpose of this study is therefore to determine the underlying process of foreign portfolio equity flows in the sub-Saharan Africa countries for which a sufficiently long data series is available (i.e., Kenya, Nigeria, South Africa, and Zambia); to establish the impact of these flows on the capitalisation of their stock markets; and draw conclusions on optimal policy choices based on this. Secondary monthly data, covering the period January 1994 to March 2019, is used, but with different sample periods for each country within that range. Structural break estimations are further undertaken to obtain more specific results. Fractal analysis is employed to estimate the Hurst parameter, a measure of the underlying process. This is aided by fractal signal classification, adopted from electronic and communication engineering and physiology, a novel approach in the analysis of capital flows, to avoid misinterpreting the estimated Hurst parameter. The correlation measure technique, another novelty in the analysis of foreign capital flows, is also used to further understand the underlying process of the flows. Bayesian techniques based on sign restrictions are employed in estimating the Calderon-Rossell model, a unique approach, to establish the impact of these flows on stock market capitalisation. The robustness of the results is tested with the Fry and Pagan Median target method. The results indicate that the underlying process of gross foreign portfolio equity inflows and outflows in the four sub-Saharan Africa countries are anti-persistent. Further, increases in market capitalisations owing to positive shocks to foreign portfolio equity inflows are greater than declines resulting from shocks to outflows. The policy implication of these results for the four SSA countries is that capital controls on foreign portfolio equity flows are redundant.

South Africa remains reliant on a number of countries to sustain its energy requirements. The acute shortage and unreliable supply of electricity, requires that South Africa consider other energy sources, more specifically, that it considers the use of oil and gas as an alternative or complementary energy source to the main energy sources currently used in the country. The recent announcement of oil and gas discoveries in South Africa could see less reliance on other countries for the importation of crude oil and petroleum products. Despite newly discovered oil and gas resources in South Africa, the country will continue to remain reliant on the industry's international investors since South Africa does not have the requisite expertise, skill and capital to operate efficiently in this industry. The shortage of capital to further develop the industry means that the country will need to continue to compete with other emerging markets to secure international investment. One such way of being an attractive investment destination has been touted through the offering of a tax regime with incentives which, more importantly, provides certainty and stability (Mausling, 2017: iv). South Africa introduced the Tenth Schedule to the Income Tax Act No.58 of 1962 (“Income Tax Act”) which aims to provide incentives that stimulate industrial and economic growth in the oil and gas industry. Against the backdrop of recent oil and gas finds, the Minister of Finance announced in the 2019 Budget Speech that changes to the Tenth Schedule would be considered. This has brought about the need to consider the role of fiscal stability agreements (“FSA”). Under the Tenth Schedule, South Africa offers FSAs which allows for the “freezing” of the Income tax provisions when the FSA was signed i.e. even if there is a new tax regime, the investor may elect to continue to use the old tax regime. Firstly, this dissertation considers whether there is a need for FSAs. To achieve this aim, the dissertation considers the reasons for fiscal instability and considers these within the South African context. These reasons are used to substantiate whether there is a need for FSA's as a remedy for fiscal instability. The current incentives offered by the Tenth Schedules are examined in order to determine the reasons as to why oil and gas companies would find FSAs advantageous. Secondly, the dissertation examines the types of FSAs typically offered, including freezing clauses (currently used in the FSA offered by South Africa), economic balancing clauses and, finally, hybrid clauses. In critically reviewing these different clauses, the most preferred clause is suggested. A further review of this preferred clause is enhanced through the consideration of the types of FSAs offered by comparable countries. Ghana and Mozambique have been identified and selected for comparison for the purposes of this study. The paper further considers aspects of the FSA such as the legality and legal effectiveness of FSAs. Such issues are critical in light of the challenges that have been identified in the use of FSAs, particularly that such instruments limit the State's sovereignty. Furthermore, the costs associated with FSAs are considered. Lastly, the remedies available should the state elect not to adhere to a FSA once in force are considered. The findings of the study suggest that there remains a need for FSAs in South Africa. However, the findings indicate a need to change the current fiscal stability clause into an economic balancing clause, in particular a negotiated economic balancing clause.

Modern portfolio theory (MPT), asset pricing models and broader financial modelling are dependent upon the accuracy of input parameters. For example, the accuracy of expected returns, standard deviations and correlations as an input into MPT will result in a more efficient selection of the optimal portfolio. These metrics are exposed to reference-day risk which is the variation in input estimation due to the selection of initial reference-day in calculations. This paper examines whether a change in reference-day, the day on which a metric is calculated, significantly affects estimates of risk-return metrics on the Johannesburg Stock Exchange (JSE). Thereafter, it applies these findings to the asset allocation problem of constructing a maximum Sharpe portfolio. The objective of this paper is to further prior research through the evaluation of an alternative simulation method and an extension of the range of tested metrics. The advancement of this prior research is achieved through the use of the Cholesky decomposition and a nonparametric bootstrapping procedure to generate reference-day risk-free estimates for average returns, standard deviations, correlations and betas. Furthermore, this paper applies the reference-day risk-free metrics to the construction of optimal multi-asset portfolios in the mean-variance framework. The findings suggest that through the use of a five-year period of monthly returns, the selection of a reference-day materially affects risk-return metrics and the subsequent portfolio characteristics that are based upon these metrics. The performance of portfolios, optimised on each reference day, ranged between 10% during the out-of-sample period. Additionally, using traditional end of month data resulted in underperformance of out-of-sample, overstated average returns, understated standard deviations and lower correlations between asset classes. Based on these findings we propose an alternative bootstrapping method for calculating reference-day risk-free metrics which reduces the effect of reference-day risk. The purpose of this methodology is to use these estimates for portfolio construction, risk management and asset pricing. The results of this paper indicate that reference-day risk makes a material difference in portfolio construction.